Virtual power plant edge control method

Abstract

The invention relates to the technical field of virtual power plant control, in particular to an edge control method of a virtual power plant, which comprises the following steps: acquiring a load power parameter matching threshold interval to generate a response level, matching a strategy template to generate a control instruction, extracting a link time delay division level to determine a synchronization reference, and analyzing the link jitter rate, screening an optimal link generation identifier, and adjusting a slave link time sequence generation time synchronization protocol. According to the invention, the active power and the power factor are collected to establish an overlapping judgment mechanism, the two-dimensional recognition of the load state and the energy efficiency level is realized, the precision and the dynamic adaptability of response grade division are improved, and the pertinence and the real-time performance of instruction generation are enhanced in combination with an adjustment threshold, a time slice and a priority matching control strategy. A main link is screened according to link time delay and jitter characteristics, the stability of a control path is guaranteed, a receiving time sequence and feedback response are corrected through time delay deviation between the main link and the slave link, and synchronous control over the signal execution process is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual power plant control, and in particular, to an edge control method for a virtual power plant. Background Art

[0002] The technical field of virtual power plant control includes technical methods for integrating, scheduling, and optimizing the control of distributed energy resources. The core content of this technical field lies in integrating multiple geographically dispersed energy units, such as photovoltaic power generation, wind power generation, energy storage systems, and adjustable loads, into a unified and coordinated virtual entity through information and communication technology and automation control technology, so as to achieve the ability to provide auxiliary services to the power system and participate in market operations. The overall technical system of virtual power plant control mainly includes a resource aggregation mechanism, a load forecasting and optimal scheduling strategy, an information interaction protocol, an energy management process, a real-time status monitoring and control execution mechanism, etc. Its system structure usually relies on a multi-level control architecture and a distributed intelligent control mechanism to ensure the coordination, responsiveness, and economy of the virtual power plant under different operating scenarios.

[0003] Among them, an edge control method for a virtual power plant refers to a technical method for independently controlling and coordinately scheduling the power resources on the edge side of the virtual power plant. It covers local measurement, status identification, and behavior prediction of controllable resources located at the end of the distribution network, such as distributed photovoltaic units, household energy storage devices, and interruptible loads, and on this basis, realizes the formulation and execution of local control decisions through a local computing terminal. The control strategy is optimized and matched according to the operating characteristic parameters of the equipment, power quality constraint conditions, and power regulation capabilities, and combines a multi-level communication protocol between the edge side and the main control center to achieve a linkage mechanism for information reporting and control response, so as to complete the integration and scheduling of edge resources in a locally autonomous manner.

[0004] The existing virtual power plant control technology has obvious limitations in dealing with the diversity of edge resource states and the fluctuations of link communication delays. During the load state recognition process, devices are usually judged only based on a single power parameter or threshold, ignoring the collaborative effects of key indicators such as power factor, resulting in a rough classification of load response levels and an inability to accurately reflect the actual operating capabilities of devices, thereby affecting the accuracy of control strategy matching. During the scheduling execution phase, in the common mode, the control center lacks the ability to monitor and dynamically discriminate link delays in real time, ignoring the command transmission errors caused by differences in link stability and feedback timeliness, resulting in command delays or asynchronous responses in the regulation of edge resources. The lack of an effective link optimization and time calibration mechanism makes the control link prone to response imbalance in high-load or network fluctuation environments, affecting the coordination and consistency of the overall control system. For example, in a multi-terminal concurrent scheduling scenario, the lag of some link feedbacks will cause misalignment of execution actions or deviation of energy management, restricting the realization of the collaborative control ability and economic objectives of the virtual power plant. The above deficiencies indicate that the existing methods have shortcomings in adaptability and accuracy in dynamic perception, hierarchical response, and link coordination. Summary of the Invention

[0005] The object of the present invention is to solve the deficiencies existing in the prior art and propose an edge control method for a virtual power plant.

[0006] To achieve the above object, the present invention adopts the following technical solutions: An edge control method for a virtual power plant, comprising the following steps: S1: Collect the current active power value, rated power value, and current power factor of the adjustable load, match the ratio of the active power value to the rated power value with a preset active power threshold interval, establish a cross-judgment relationship in combination with the power factor interval classification of the power factor, and generate an overlapping area load response level identifier; S2: Based on the overlapping area load response level identifier, match the corresponding level's active power adjustment threshold, time slice scheduling strategy, and control action priority, and generate an edge control instruction set in combination with the target power adjustment requirement; S3: Combine the time slice scheduling strategy in the edge control instruction set, obtain the scheduling command sending timestamp and feedback response timestamp of the control link in real time, extract the link round-trip delay and response duration, sort the delay values to divide the delay level interval and determine the maximum delay value, and generate a synchronous time window reference value; S4: According to the synchronous time window reference value, analyze the delay jitter rate of the candidate main link within the control period, screen the candidate links that meet the requirements in combination with the delay stability threshold, select the optimal link according to the delay value ranking, and generate a main link optimization identifier.

[0007] As a further solution of the present invention, the overlapping area load response level identifier includes active power threshold interval classification, power factor interval classification, and response level mapping relationship. The edge control instruction set includes active power adjustment threshold, time slice scheduling strategy, and control action priority. The synchronous time window reference value includes control link round-trip delay, response duration, and maximum delay value. The main link preference identifier includes delay jitter rate, delay stability threshold, and optimal link sorting result.

[0008] As a further solution of the present invention, the specific steps of S1 are as follows: S101: Collect the current active power value, rated power value, and current power factor of the adjustable load, record the original values of the active power value and the rated power value, classify the power factor into the preset power factor interval classification, and generate a real-time power parameter set. S102: Based on the real-time power parameter set, calculate the ratio of the active power value to the rated power value, perform range matching on the ratio with the preset active power threshold interval, determine the threshold interval number where the ratio is located, and generate an active power ratio interval identifier. S103: Call the active power ratio interval identifier and the power factor interval classification, establish a cross-judgment table of the ratio interval number and the power factor classification, generate a classification code according to the overlapping area corresponding rule, and generate an overlapping area load response level identifier.

[0009] As a further solution of the present invention, the specific steps of S2 are as follows: S201: Based on the overlapping area load response level identifier, match the active power adjustment threshold, time slice scheduling strategy, and control action priority corresponding to the level in the strategy template library, and generate a load response parameter group by matching the numerical range of the load response level identifier with the template library threshold interval. S202: Call the time slice scheduling strategy in the load response parameter group, divide the total period into time slices according to the deviation amount of the target power adjustment requirement and the time constraint according to the preset rule, and adjust the time slice length and allocation ratio in combination with the deviation amount distribution to generate a time slice configuration sequence. S203: Based on the time nodes and allocation ratio of the time slice configuration sequence, combine the adjustment threshold of the load response parameter group and the priority sorting rule, and map the deviation amount to the time slice according to the priority to generate an edge control instruction set.

[0010] As a further solution of the present invention, the specific calculation formula for mapping the deviation amount to the time slice according to the priority is: ; Wherein, represents the deviation allocation amount of the th time slice. represents the priority weight of the th time slice, where wj is the weight coefficient of the jth time slice, represents the absolute value of the total system load deviation, represents the allocation ratio of the th time slice, represents the total number of time slices, represents the adjustment threshold correction coefficient, represents the adjustment threshold of the th time slice.

[0011] As a further solution of the present invention, the specific steps of S3 are as follows: S301: Based on the time slice scheduling strategy in the edge control instruction set, extract the sending timestamp and feedback response timestamp of the scheduling command, perform corresponding matching according to the control command number, establish a timestamp pair and form a link scheduling information set, calculate the round-trip delay between commands based on the timestamp pair, and generate a link round-trip delay value sequence after summarization; S302: Call the link round-trip delay value sequence, arrange it in the order of delay values, divide continuous level intervals, and perform boundary correction according to the distribution characteristics of the interval differences to obtain the link delay level division interval; S303: According to the link delay level division interval, extract the maximum round-trip delay value in the highest interval, determine the synchronous time domain boundary in combination with the scheduling period parameter, and generate the synchronous time window reference value.

[0012] As a further solution of the present invention, the specific steps of S4 are as follows: S401: Based on the synchronous time window reference value, collect the delay sequence of the candidate primary link within the control period, calculate the delay jitter rate of the link within the period according to the change amplitude and frequency between adjacent delay values, and generate the candidate link delay jitter rate; S402: According to the candidate link delay jitter rate, compare it with the delay stability threshold under the synchronous time window reference value, extract the average delay value of the retained links for sorting, and obtain the stability screening sorting sequence value; S403: Based on the stability screening sorting sequence value, analyze the identification information of the link ranked first and write it into the corresponding position field in the screening link set to establish the primary link preference identification.

[0013] As a further solution of the present invention, the specific calculation formula for calculating the delay jitter rate of the link within the period according to the change amplitude and frequency between adjacent delay values is: ; where Represents the link delay jitter rate, Represents the Absolute difference of adjacent delay values, Represents the Reciprocal of the delay sampling interval, Represents the Dynamic weight factor of the Represents the fixed cycle duration of the synchronization time window reference value, Represents the change frequency of adjacent delay values within the control period, Represents the frequency smoothing coefficient.

[0014] As a further solution of the present invention, the method further includes, S5: Based on the instruction issuance time of the preferred identifier of the main link, call the synchronization time window reference value, and adjust the signal reception start point and feedback timing offset of the slave link in combination with the delay deviation value between the slave link and the main link to generate the master-slave link time synchronization protocol; The master-slave link time synchronization protocol includes the instruction issuance time, the signal reception start point, and the feedback timing offset.

[0015] As a further solution of the present invention, the specific steps of S5 are, S501: Based on the instruction issuance time of the preferred identifier of the main link, call the synchronization time window reference value and compare it with the main link signal reception start time, screen the main link signals that meet the time interval as the synchronization reference, obtain the time interval range between the instruction issuance time, and generate the main link synchronization interval value; S502: According to the main link synchronization interval value, calculate the delay deviation between the link signal reception time and the main link signal reception time, judge the correction requirement of the slave link signal reception position based on the deviation, and adjust the reception time accordingly to obtain the slave link signal correction start point; S503: Based on the slave link signal correction start point, combine the main link reception time and the slave link feedback timing time to construct a feedback timing comparison interval, adjust the feedback timing offset according to the time difference and interval relationship, and establish the master-slave link time synchronization protocol.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by collecting the active power and power factor to establish an overlapping determination mechanism, the two-dimensional identification of the load state and energy efficiency level is realized, the accuracy and dynamic adaptability of the response level division are improved, combined with the adjustment threshold, time slice and priority matching control strategy, the pertinence and real-time nature of the instruction generation are enhanced, the main link is selected according to the link delay and jitter characteristics to ensure the stability of the control path, and the reception timing and feedback response are corrected through the delay deviation between the master and slave links to realize the synchronous control of the signal execution process. Brief Description of the Drawings

[0017] Figure 1 It is a schematic diagram of the step flow of the present invention; Detailed Embodiment

[0018] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0019] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, in the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined.

[0020] Please refer to Figure 1 , a method for edge control of a virtual power plant, including the following steps: S1: Collect the current active power value, rated power value and current power factor of the adjustable load, match the ratio of the active power value to the rated power value with a preset active power threshold interval, establish a cross-judgment relationship by classifying the power factor intervals of the power factor, and generate an overlapping area load response level identifier; Active power threshold interval: The allowable fluctuation range of active power set according to the power system dispatching specification; Power factor interval classification: The power factor range divided based on the power industry standard (such as IEEE1547); S2: Based on the overlapping area load response level identifier, match the corresponding level of active power adjustment threshold, time slice scheduling strategy and control action priority from the policy template library, and generate an edge control instruction set in combination with the target power adjustment requirement; Active power adjustment threshold: The upper and lower limits of power adjustment allowed by the power equipment.

[0021] Time slice scheduling strategy: The instruction execution time unit divided based on the real-time control requirement; S3: Based on the time slice scheduling strategy in the edge control instruction set, obtain the scheduling command sending timestamp and feedback response timestamp of the control link in real time, extract the link round-trip delay and response duration, sort according to the delay value to divide the delay level interval and determine the maximum delay value; generate the synchronous time window reference value; Round-trip delay: The total time difference from the instruction sending to the feedback receiving in the communication link.

[0022] Delay level interval: The delay segmentation defined according to the communication protocol (such as IEC61850); Synchronous time window reference value: The maximum allowable round-trip delay value within the control period, used to constrain the timing range of the communication link collaborative operation; S4: Based on the synchronous time window reference value, analyze the delay jitter rate of the candidate master link within the control period, combine with the delay stability threshold to screen the candidate links that meet the requirements, and select the optimal link according to the delay value sorting; generate the master link preference identifier; Delay jitter rate: The ratio of the standard deviation of the link delay fluctuation to the average delay; Delay stability threshold: The maximum allowable jitter rate set according to the communication reliability requirements; S5: Based on the instruction issuing time of the master link preference identifier, call the synchronous time window reference value, and adjust the signal receiving starting point and feedback timing offset of the slave link in combination with the delay deviation value between the slave link and the master link; generate the master-slave link time synchronization protocol.

[0023] Delay deviation value: The difference between the round-trip delays of the slave link and the master link; Master-slave link time synchronization protocol: The timing calibration rule based on IEEE1588 (PTP protocol), used for multi-link collaborative control.

[0024] The overlapping area load response level identifier includes the active power threshold interval classification, power factor interval classification, and response level mapping relationship. The edge control instruction set includes the active power adjustment threshold, time slice scheduling strategy, and control action priority. The synchronous time window reference value includes the control link round-trip delay, response duration, and maximum delay value. The master link preference identifier includes the delay jitter rate, delay stability threshold, and optimal link sorting result. The master-slave link time synchronization protocol includes the instruction issuing time, signal receiving starting point, and feedback timing offset.

[0025] The specific steps of S1 are as follows: S101: Collect the current active power value, rated power value, and current power factor of the adjustable load, record the original values of the active power value and the rated power value, classify the power factor into the preset power factor interval classification, and generate the real-time power parameter set; When collecting real-time adjustable load data on-site, it is necessary to continuously obtain the real-time working state parameters of the target device through intelligent measurement and control devices installed in the power distribution system, including the active power output at present, the rated power indicated on the equipment nameplate, and the current power factor. Among them, the acquisition of active power can rely on measuring the instantaneous values of voltage and current and combining the phase angle difference between voltage and current to obtain the actual power output. The rated power parameter can be directly retrieved from the database of the equipment management platform. Usually, it is checked according to the factory data and the engineering configuration list during input to ensure accuracy. The power factor needs to detect the phase relationship of the voltage and current waveforms through the power analysis unit in the device to obtain the effective conversion ratio of the current electric energy. For three-phase equipment, three-phase data needs to be collected for averaging or characteristic values are selected according to the main load phase. For example, in the air-conditioning system of an office building, the voltage value is 220 volts, the current is 4.5 amperes, the measured power factor is 0.85, the corresponding calculated actual active power is 841.5 watts, and the equipment rated power is 1000 watts. Then 841.5 watts and 1000 watts are used as the operation state records of this equipment at this moment, and they are classified according to the power factor range. The set power factor classification ranges include, for example, [0, 0.6), [0.6, 0.8), [0.8, 0.9), [0.9, 1], etc. The current 0.85 belongs to the third gear. The finally recorded set of real-time power parameters includes three parts: the current active power value, the rated power value, and the power factor range, forming a structured record for subsequent judgment and processing.

[0026] S102: Based on the set of real-time power parameters, calculate the ratio of the active power value to the rated power value, match the ratio with the preset active power threshold range, determine the threshold range number where the ratio is located, and generate an active power ratio range identifier; Compare the currently acquired active power value with the corresponding rated power value in real time. First, calculate the ratio based on the actual values by dividing the current output by the nameplate rated output to obtain the proportional relationship. This ratio represents the relationship between the load operation state and the design capacity. To facilitate determination and classification, match this ratio with a pre-set threshold interval table, usually divided into several operation level intervals, such as setting the intervals as [0, 0.2), [0.2, 0.4), [0.4, 0.6), [0.6, 0.8), [0.8, 1.0], etc. Taking the previously acquired equipment as an example, the current active power is 841.5 watts and the rated power is 1000 watts, with a ratio of 0.8415, which falls into the [0.8, 1.0] interval and corresponds to the fourth gear number. During this process, ensure that all ratio calculations are based on valid data. If the rated power item is missing or zero, this item does not participate in the comparison. The threshold intervals are set through statistical classification based on a large amount of load operation data and are reasonably set according to the distribution of power utilization rates of equipment during actual operation to avoid being too dense or too sparse and ensure that the classification is distinguishable and applicable. The ratio attribution of each equipment corresponds to a number one by one, and finally output this number as the identification information of the power ratio interval, which is used as one of the subsequent classification bases.

[0027] S103: Call the active power ratio interval identification and power factor interval classification, establish a cross-judgment table of the ratio interval number and power factor classification, generate a classification code according to the corresponding rules of the overlapping area, and generate a load response level identification for the overlapping area; Use the power ratio interval number obtained above and the classification of the power factor interval as dual indicators to construct a cross-judgment matrix, form a two-dimensional classification table, set the power factor interval horizontally and the ratio interval number vertically, and fill in the corresponding classification code in each cross unit. This code reflects the classification level of the current load under specific conditions. The code is designed according to the load characteristics and the grading standard of the regulation response ability. For example, class A represents low regulation ability, class D represents high regulation ability, and the numbers represent further sub-grades. For example, if the power ratio number is the fourth gear and the power factor interval belongs to [0.8, 0.9), the corresponding item in the cross-table is set to "D3", indicating that the equipment is in a relatively high operation state and the power factor is in a relatively optimal interval. The classification number setting process constructs a rule base by combining a large number of historical operation state samples, and is automatically generated by the system according to the set rules. The classification table is called by the data processing module. Whenever new real-time data is received, it immediately performs matching and code extraction to determine the response level where the load is currently located in the overlapping area, and outputs the corresponding level identification as an input item for subsequent control and analysis.

[0028] The specific steps of S2 are as follows: S201: Based on the load response level identifier of the overlapping area, match the active power adjustment threshold, time slice scheduling strategy, and control action priority corresponding to the level in the strategy template library. Generate a load response parameter group by matching the numerical range of the load response level identifier with the threshold interval in the template library; The load response level identifier of the overlapping area is jointly determined by three factors, namely historical response ability, current response status, and geographical location factor. Suppose the historical response ability of a certain load unit is 85%, the current response status is 75%, and the geographical location factor is 50%. The response level obtained after weighted combination is 0.74. The response level is in the interval of 0.7 to 0.8, and the strategy template in this level interval in the strategy template library is matched. The active power adjustment threshold specified by this template is 500 kW, the time adjustment period is 15 minutes, and the control action priority is that load A is prioritized, followed by load C, and then load B. The system matches the actual response level with the level interval of the strategy template. After confirmation of the match, the above three key parameters are read from the template. The finally generated load response parameter group includes: the adjustment power is 500 kW, the adjustment duration is 15 minutes, and the response control order is three load units A, C, and B. This parameter group will be used as the basis for subsequent adjustment and instruction generation to ensure that the allocation of adjustment resources is logically dispatched based on the response level.

[0029] S202: Invoke the time slice scheduling strategy in the load response parameter group. Divide the total cycle into time slices according to the preset rules based on the deviation amount of the target power adjustment demand and the time constraint, and adjust the time slice length and allocation ratio in combination with the deviation amount distribution to generate a time slice configuration sequence; The scheduling strategy first calls the adjustment cycle information in the parameter group, that is, 15 minutes as the total time. If the target power adjustment requirement is 450 kW, exceeding the allowable error range, for example, a tolerance deviation of 5% which is 22.5 kW, then this 15-minute cycle needs to be divided into multiple time slices. Initially, the cycle is evenly divided into three time slices, each being 5 minutes. Then, according to the distribution of the past load response capabilities, the adjustment targets are allocated proportionally. If historical data shows that the response capability of the first segment accounts for 50%, the second segment for 30%, and the third segment for 20%, then these three time slices are respectively assigned adjustment tasks of 225, 135, and 90 kW. After that, by combining the response speeds of each time slice, it is evaluated that the average response rate of the first segment is 90%, the second segment is 60%, and the third segment is 50%. Through the analysis of the relationship between the response speed and the desired adjustment power, it can be seen that the actual response time required for each segment changes. After adjustment, the first segment is about 5.6 minutes, the second segment is about 6.25 minutes, and the third segment is about 9 minutes, which exceeds the total cycle. Therefore, compression adjustment is required to make the sum of the three segments equal to 15 minutes again. After compression, the three time slices are finally re-allocated to 4.3 minutes, 5.2 minutes, and 5.5 minutes. This configuration is used to guide the load response rhythm at different stages and form a clear scheduling beat.

[0030] S203: Based on the time nodes and allocation ratios of the time slice configuration sequence, combined with the adjustment thresholds and priority sorting rules of the load response parameter group, map the deviation amount to the time slices according to the priority to generate the edge control instruction set; The specific calculation formula for mapping the deviation amount to the time slices according to the priority is: ; Among them, represents the deviation allocation amount of the th time slice, represents the priority weight of the th time slice, wj is the weight coefficient of the jth time slice, represents the absolute value of the total system load deviation amount, represents the th time slice's allocation ratio, is the allocation ratio of the jth time slice, represents the total number of time slices, represents the adjustment threshold correction coefficient, represents the th time slice's adjustment threshold; Detailed explanation of the formula and the derivation process of the formula calculation: The absolute value of the total system load deviation amount is obtained by collecting the difference between the current total load and the target value in real time through the power grid monitoring system and taking the absolute value. In the monitoring data of a certain regional power grid , priority weight Generated according to the preset rules of the load response parameter group, time slice weight , allocation ratio Calculated from historical load distribution data, time slice allocation ratio , adjustment threshold correction coefficient Set according to the device adjustment ability to , time slice adjustment threshold Determined by the device response characteristics, time slice adjustment threshold , total number of time slices (Divided at 15-minute intervals based on a 24-hour cycle).

[0031] Substitute into the formula to calculate the deviation allocation amount of the time slice : : The numerator part operation is , the summation term in the denominator Calculated according to all time slice data, assuming that the average value is and average value is , then , after taking the square root , the adjustment threshold correction term is , the total denominator is , finally . This result indicates that the time slice needs to allocate approximately of the load adjustment amount to generate the edge control instruction for the corresponding time slice.

[0032] Parameter definition: is the deviation allocation amount of the th time slice, is the weight coefficient generated based on the load response priority rule for the th time slice, is the absolute value of the total load deviation collected by the power grid monitoring system in real time, is the th time slice allocation ratio calculated based on historical load distribution, is the total number of time slices, is the threshold correction coefficient set according to the device adjustment ability, is the th time slice adjustment threshold determined by the device response characteristics.

[0033] The specific steps of S3 are: S301: Based on the time slice scheduling strategy in the edge control instruction set, extract the sending timestamp and feedback response timestamp of the scheduling command, match them according to the control command number, establish a timestamp pair and form a link scheduling information set, calculate the round-trip delay between commands based on the timestamp pair, and generate a link round-trip delay value sequence after aggregation; The edge control instruction centralized time slice scheduling strategy is adopted. First, the round-robin scheduling method needs to be set in the edge node. The execution window of the control command is allocated according to the preset time slice. Each time a command is sent, the current system time is recorded as the sending timestamp. After the command is executed, the system receives the corresponding feedback response and records the receiving time again as the response timestamp. The two timestamps are matched one by one according to the number of the control command to form a timestamp pair and a scheduling link information set. For example, in an automated assembly line, the edge node records the sending time as 100.1 milliseconds when controlling the extension of the robot arm, and the feedback position has reached the response time of 100.1 milliseconds. The response time is 102.3 milliseconds, and the corresponding command number is set to 105. The matched timestamp pair is <100.1, 102.3>. The command number is used for corresponding confirmation to ensure that each pair of timestamps belongs to the same control process. Subsequently, the round-trip delay of the command is calculated to be 2.2 milliseconds through the difference between the two timestamps. In this way, timestamp pairs from different commands are continuously accumulated, and the round-trip delay of each group is calculated separately to form a group of link delay data, such as 2.2 milliseconds, 2.0 milliseconds, 2.3 milliseconds, 2.1 milliseconds, etc. This data sequence is the link round-trip delay value sequence, which becomes the basis for subsequent scheduling analysis.

[0034] S302: calling the link round-trip delay value sequence, arranging them in order of delay values, dividing them into continuous grade intervals, performing boundary correction according to the distribution characteristics of interval differences, removing deviated sections and completing interval delineation, and obtaining link delay grade division intervals; The aforementioned round-trip delay value sequence of the link is called and processed. First, all delay values ​​are arranged in ascending order to ensure a clear data distribution structure. Then, all sorted delay values ​​are divided into continuous time intervals according to the set time division unit. For example, the width of each level interval is set to 1 millisecond, and multiple intervals are generated in sequence. The number of delays contained in each interval is counted. According to the statistical results, the situation in which the number of data in some intervals is obviously small is identified, and it is identified as a deviation segment. For this purpose, a calculation method for determining density difference is introduced. By calculating the proportion of the difference in the number of data between adjacent intervals, if the proportion is greater than 80%, it is considered a density mutation, thereby Perform interval boundary correction operations. On this basis, only segments with concentrated data density are retained to form the final delay level interval. For example, in actual data, the initial interval is multiple 1-millisecond wide intervals from 2 milliseconds to 9 milliseconds, and the density is concentrated between 2 and 4 milliseconds and between 7 and 9 milliseconds. The data in several intermediate segments is sparse, so the final delineated intervals are two: 2 to 4 milliseconds and 7 to 9 milliseconds. While dividing the level intervals, the numerical features are combined to identify the segments in the high delay range. For example, when the minimum value of an interval exceeds the mean value by more than two standard deviations, it can be determined that the interval is in the significant delay level, ensuring that the boundary delineation reflects the performance fluctuation characteristics of the actual network link.

[0035] S303: Divide the intervals according to the link delay level, extract the maximum round-trip delay value in the highest interval, determine the synchronization time domain boundary in combination with the scheduling cycle parameters, and generate a synchronization time window reference value; According to the link delay level interval obtained in the previous processing process, the item with the highest value in the maximum level interval is selected as the representative indicator. The maximum value comes from the highest round-trip delay of all successfully matched timestamp pairs, indicating the ultimate delay performance that the current link may reach. For example, the maximum level interval is seven to nine milliseconds, and the maximum round-trip delay is eight point one milliseconds. In the actual scheduling environment, such as the intelligent warehousing system for robot path adjustment control, the system can use this maximum round-trip delay as the starting point of the scheduling time domain boundary, combined with the set scheduling cycle parameters, such as ten milliseconds, through a simple numerical addition method, add a fixed delay compensation value to generate a reference value of the synchronization time window. For example, if the compensation value is set to 0.5 milliseconds, the reference value is 8.6 milliseconds. In the subsequent synchronization scheduling process, the system will construct the synchronization time window according to this reference value. For example, if the tolerance range is set to plus or minus 0.2 milliseconds, the final synchronization time domain judgment range is 8.4 to 8.8 milliseconds. This range is used to determine whether each edge node meets the unified time domain requirements to ensure that the scheduling pace can still be unified under the influence of the maximum link delay.

[0036] The specific steps of S4 are: S401: Based on the synchronization time window reference value, collect the delay sequence of the candidate primary link within the control period. Calculate the delay jitter rate of the link within the period according to the change amplitude and frequency between adjacent delay values, and generate the candidate link delay jitter rate. The specific calculation formula for calculating the delay jitter rate of the link within the period according to the change amplitude and frequency between adjacent delay values is: ; Among them, represents the link delay jitter rate, represents the absolute difference between adjacent delay values, represents the reciprocal of the delay sampling interval, represents the dynamic weight factor of the th delay difference, represents the fixed cycle duration of the synchronization time window reference value, represents the change frequency between adjacent delay values within the control period, represents the frequency smoothing coefficient ( ); Detailed explanation of the formula and the derivation process of formula calculation: Parameter assignment and acquisition method: , , , , , a total of delay values, and the change frequency of adjacent delays .

[0037] Absolute difference between adjacent delays: , , , .

[0038] Reciprocal of the sampling interval: According to the sampling frequency of the network monitoring device (interval ), .

[0039] Synchronization time window reference value: Set , based on the typical value of the synchronization time window in the network protocol standard (refer to IEEE802.1AS).

[0040] Frequency smoothing coefficient: Set , according to the empirical formula , when , round down to 。

[0041] Dynamic weight factor: Calculate , where , , to obtain ; similarly , , 。

[0042] Step - by - step calculation of the formula: Calculation of the numerator: ; ; ; ; Sum of the numerators ; Calculation of the denominator: ; Final result: ; Explanation of the numerical result: The calculation result is the link delay jitter rate, and its value reflects the comprehensive fluctuation degree of the delay change within the control period. The larger the value, the more significant the delay jitter.

[0043] This result, through normalization processing (the denominator contains and the frequency term) and the introduction of the dynamic weight factor , correlates parameters such as the amplitude of adjacent delay differences, change frequency, and sampling density, and is used to horizontally compare the jitter stability of different candidate links, directly related to the generation of the candidate link delay jitter rate at the end of the step.

[0044] S402: According to the candidate link delay jitter rate, compare with the delay stability threshold under the synchronous time window reference value, screen out the links whose jitter rate exceeds the threshold, extract the average delay values of the remaining links for sorting, and obtain the stability screening and sorting sequence value; After obtaining the delay jitter rate of each link, a threshold for distinguishing stable and unstable states is set. For example, the threshold is set to twenty percent. All links with jitter rates higher than this threshold are regarded as unstable and are removed. Only the links with jitter rates not exceeding the threshold are retained. For each retained link, the average delay value within its entire cycle is calculated, and then these values are sorted in ascending order to form a sorted sequence. The specific process includes: for each retained link, add up all the delay data from start to end and divide by the total amount of data to obtain the average delay of this link. Then, compare the average delay values of all links and generate a sorted list in ascending order. For example, for the two retained links with cycle delays of 10, 10, 10, 10, 10 milliseconds and 10, 10, 11, 10, 10 milliseconds respectively, the corresponding average values are 10 milliseconds and 10.2 milliseconds. After sorting, the former is ranked first, and the acquisition of the stability screening sorted sequence is completed.

[0045] S403: Based on the values of the stability screening sorted sequence, analyze the identification information of the link ranked first and write it into the corresponding position field in the screened link set to establish the preferred identification of the main link; Analyze the link sequence after stability sorting, take out the link ranked first. Its identification information is used as the current main link. Find the corresponding link position in the established link information set, and assign the main link identification field of this link to the main link state. The specific operation is to read the link number of the first item in the sorting result, find its corresponding information unit in the link data structure, and set the boolean or integer field used to identify the main link state to the active state, while the remaining link fields remain in the inactive state. Thus, the operation of assigning the preferred identification of the main link is completed.

[0046] The specific steps of S5 are as follows: S501: Based on the instruction issuance time of the preferred identification of the main link, call the synchronous time window reference value and compare it with the start time of receiving the main link signal to screen the main link signal that meets the time interval as the synchronization reference, obtain the time interval range between the instruction issuance time, and generate the main link synchronization interval value; In the main link, it is necessary to extract the instruction issuance time from the data record, the reception start time of each frame of signal, and the reference value of the synchronization time window set in the system. Compare one by one the intervals between the reception times of all main link signals and the instruction issuance time, and filter out those signal frames whose time intervals are within the set window range as the reference basis for subsequent synchronization operations. For example, if the set reference value is 5 milliseconds, the instruction issuance time is 200 milliseconds, and the reception start times of the signals are 195 milliseconds, 198 milliseconds, 202 milliseconds, and 206 milliseconds, then only 198 milliseconds and 202 milliseconds meet the condition within the set range. Based on this, a synchronization interval is established, ranging from 198 milliseconds to 202 milliseconds, and this time period is the main link synchronization interval. During this process, the time stamp recording module should be used to accurately read the time information of the signal frames, and the optimal synchronization comparison samples should be extracted according to the time offset constraint, and finally the reference interval for time alignment between the main link and the slave link can be obtained. Such a scheme can be applied to multi-channel signal processing devices, such as distributed radar networks. After a synchronization control instruction is issued at a certain node, the remaining nodes locate the legal time window for signal reception according to this instruction, so as to select the accurate time alignment reference point.

[0047] S502: Calculate the delay deviation between the reception time of the link signal and the reception time of the main link signal according to the main link synchronization interval value, judge the correction requirement of the reception position of the slave link signal according to the deviation, and adjust the reception time accordingly to obtain the correction start point of the slave link signal; According to the synchronization time interval determined by the main link, compare the times of the received signals in the slave link correspondingly, and analyze the time difference between the reception time of the slave link and the reception time of the main link signal. If the time interval between the reception time of a certain slave link signal and the time of the main link signal exceeds the set maximum allowable deviation value, it can be judged that the signal reception point has a too large offset and needs to be corrected. At this time, the reception time should be adjusted to reduce the offset time within the main link reference time range. For example, the main link signal is 200 milliseconds, the slave link signal is 204 milliseconds, and the set maximum deviation is 2 milliseconds, then the current deviation is 4 milliseconds, which has exceeded the tolerance range. The time of the slave link signal should be adjusted to be consistent with the main link to serve as the correction start point. If the deviation is negative, reverse correction is performed. After correction, the offset trend needs to be continuously tracked, and the adjustment value required for each correction should be recorded. In practical applications, in large communication systems or distributed sensor platforms, there are slight jitters in the signal reception of each sub-node. The correction process is adjusted by comparing the main link reference value to ensure that all signals can be aligned within a unified time reference.

[0048] S503: Based on the correction starting point of the slave link signal, combine the master link reception time and the slave link feedback timing time to construct a feedback timing comparison interval, adjust the feedback timing offset according to the relationship between the time difference and the interval, and establish a master-slave link time synchronization protocol; Combine the correction starting point of the slave link, the master link reception time, and the slave link feedback time to construct a comparison interval between the feedback timings. According to the time difference within this interval, determine whether there is an offset in the feedback timing. If the offset exceeds the set value, correct the feedback time to make it closer to the master link timing to ensure that the feedback behavior is consistent with the master link instruction. For example, if the master link reception time is 200 milliseconds, the slave link feedback time is 207 milliseconds, and the preset allowable feedback error is 5 milliseconds, then the current feedback time offset exceeds the set range and should be adjusted to meet the error tolerance, such as adjusted to 205 milliseconds. Next, construct the time synchronization protocol parameters for the master-slave link, including the feedback timing offset limit value, the master-slave reception time error range, the time offset correction term, etc. By analyzing the time data before and after correction, set the maximum time offset between the master and slave not to exceed a certain tolerance, such as set to 2 milliseconds, to ensure that the data return is consistent with the master link time reference. This set of suitable synchronization protocol parameters is applicable to an automated control system, such as a distributed control module in rail transit. Each control unit feeds back according to the timing control signal sent by the master link and corrects its own time identifier to ensure that the system actions are coordinated.

[0049] The above is only the preferred embodiment of the present invention and does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A virtual power plant edge control method, characterized in that: The following steps are involved: S1: Collect the current active power value, rated power value and current power factor of the adjustable load, match the ratio of the active power value to the rated power value with the preset active power threshold range, establish a cross judgment relationship in combination with the power factor range classification of the power factor, and generate an overlap area load response level identification; S2: Based on the overlapped area load response level identifier, matching the corresponding level of active power adjustment threshold, time slice scheduling strategy and control action priority, and combining the target power adjustment requirements, generating an edge control instruction set; S3: Combined with the time slice scheduling strategy in the edge control instruction set, the scheduling command sending timestamp and feedback response timestamp of the control link are obtained in real time, the link round-trip delay and response duration are extracted, the delay level interval is divided according to the delay value sorting and the maximum delay value is determined, and the synchronization time window reference value is generated; S4: Analyze the delay jitter rate of the candidate main link within the control period according to the synchronization time window reference value, select the candidate links that meet the requirements in combination with the delay stability threshold, select the optimal link by delay value, and generate a main link optimization identifier.

2. The edge control method of a virtual power plant according to claim 1, characterized in that: The overlapping area load response level identification includes active power threshold interval classification, power factor interval classification, and response level mapping relationship; the edge control instruction set includes active power adjustment threshold, time slice scheduling strategy, and control action priority; the synchronization time window reference value includes control link round-trip delay, response duration, and maximum delay value; the main link preferred identification includes delay jitter rate, delay stability threshold, and optimal link sorting result.

3. The edge control method of a virtual power plant according to claim 1, characterized in that: The specific steps of S1 are: S101: collecting the current active power value, rated power value and current power factor of the adjustable load, recording the original values ​​of the active power value and the rated power value, classifying the power factor into a preset power factor interval classification, and generating a real-time power parameter set; S102: Based on the real-time power parameter set, calculate the ratio of the active power value to the rated power value, match the ratio with a preset active power threshold interval, determine the threshold interval number where the ratio is located, and generate an active power ratio interval identifier; S103: calling the active power ratio interval identifier and the power factor interval classification, establishing a cross determination table of the ratio interval number and the power factor classification, generating a classification code according to the overlapping area corresponding rule, and generating an overlapping area load response level identifier.

4. The edge control method of a virtual power plant according to claim 1, characterized in that: The specific steps of S2 are: S201: Based on the overlapped area load response level identifier, matching the active power adjustment threshold, time slice scheduling strategy and control action priority of the corresponding level in the strategy template library, and generating a load response parameter group by matching the value range of the load response level identifier with the template library threshold range; S202: calling the time slice scheduling strategy in the load response parameter group, dividing the total cycle into time slices according to the deviation amount and time constraint of the target power adjustment demand according to the preset rules, adjusting the time slice length and allocation ratio in combination with the deviation amount distribution, and generating a time slice configuration sequence; S203: Based on the time nodes and allocation ratios of the time slice configuration sequence, combined with the adjustment threshold and priority sorting rules of the load response parameter group, the deviation amount is mapped to the time slice according to the priority, and an edge control instruction set is generated.

5. The edge control method of a virtual power plant according to claim 4, characterized in that: The specific calculation formula for mapping the deviation amount to the time slice according to the priority is: ； in, Representative The deviation allocation of time slices, Representative The priority weight of a time slice, w j is the weight coefficient of the jth time slice, Represents the absolute value of the total load deviation of the system, Representative The allocation ratio of time slices, is the allocation ratio of the jth time slice, Represents the total number of time slices, represents the adjustment threshold correction factor, Representative The adjustment threshold for each time slice.

6. The edge control method of a virtual power plant according to claim 1, characterized in that: The specific steps of S3 are: S301: Based on the time slice scheduling strategy in the edge control instruction set, extract the sending timestamp and feedback response timestamp of the scheduling command, perform corresponding matching according to the control command number, establish a timestamp pair and form a link scheduling information set, calculate the round-trip delay between commands based on the timestamp pair, and generate a link round-trip delay value sequence after aggregation; S302: calling the link round-trip delay value sequence, arranging them in order of delay values, dividing them into continuous grade intervals, and performing boundary correction according to the distribution characteristics of interval differences to obtain link delay grade division intervals; S303: Divide the intervals according to the link delay level, extract the maximum round-trip delay value in the highest interval, determine the synchronization time domain boundary in combination with the scheduling cycle parameters, and generate a synchronization time window reference value.

7. The edge control method of a virtual power plant according to claim 6, characterized in that: The specific steps of S4 are: S401: Based on the synchronization time window reference value, collect the delay sequence of the candidate main link within the control period, calculate the delay jitter rate of the link within the period according to the change amplitude and frequency between adjacent delay values, and generate the delay jitter rate of the candidate link; S402: extracting average delay values ​​of the reserved links according to the delay jitter rate of the candidate links and comparing them with the delay stability threshold under the synchronization time window reference value, and sorting them to obtain a stability screening sorting sequence value; S403: Filter the sorting sequence value according to the stability, analyze the identification information of the first-ranked link and write it into the corresponding position field in the filtered link set, and establish the primary link preferred identification.

8. The edge control method of a virtual power plant according to claim 7, characterized in that: The specific calculation formula for calculating the delay jitter rate of the link within a period according to the change amplitude and frequency between adjacent delay values ​​is: ； in, Represents the link delay jitter rate, Representative The absolute difference between adjacent delay values, Representative The inverse of the delay sampling interval, Representative The dynamic weight factor of the delay difference is Represents the fixed period duration of the synchronization time window reference value, Represents the frequency of changes in adjacent delay values ​​within the control period, Represents the frequency smoothing coefficient.

9. The edge control method of a virtual power plant according to claim 1, characterized in that: The method also include, S5: Based on the instruction issuing time of the master link preferred identifier, calling the synchronization time window reference value, adjusting the signal receiving starting point and feedback timing offset of the slave link in combination with the delay deviation value between the slave link and the master link, and generating the master-slave link time synchronization protocol; The master-slave link time synchronization protocol includes instruction issuing time, signal receiving starting point, and feedback timing offset.

10. The edge control method of a virtual power plant according to claim 9, characterized in that: The specific steps of S5 are: S501: Based on the instruction issuing time of the primary link preferred identifier, call the synchronization time window reference value and compare it with the primary link signal receiving start time, select the primary link signal that meets the time interval as a synchronization reference, obtain the time interval range between the instruction issuing time, and generate the primary link synchronization interval value; S502: Calculate the delay deviation between the link signal reception time and the main link signal reception time according to the main link synchronization interval value, determine the correction requirement of the slave link signal reception position according to the deviation, adjust the reception time accordingly, and obtain the correction starting point of the slave link signal; S503: Based on the slave link signal correction starting point, combined with the master link reception time and the slave link feedback timing time, a feedback timing comparison interval is constructed, the feedback timing offset is adjusted according to the relationship between the time difference and the interval, and a master-slave link time synchronization protocol is established.

Images